The goals of the proposed studies are to determine the properties of central visual mechanisms that are sensitive to colored patterns and to place this knowledge in the context of natural viewing situations, in which the visual system is exposed to lights reflected from surfaces. The proposed work aims to discover how the visual system constructs a constant, stable world populated by objects with constant shape and color, with properties that do not depend on the vagaries of scene illumination. Central visual mechanisms operate on a reflected light stimulus to separate surface properties from illumination properties and so determine the constant shapes and colors of objects. Four projects on the appearance and detectability of color patterns of a computer-driven visual display are proposed to measure and model these mechanisms. The first project concerns the way in which the color of a surface is influenced by the properties of immediately surrounding surfaces. Experiments are proposed on the appearance of time-varying patterns with the aim of characterizing mechanisms that mediate such local influence, including mechanisms of simultaneous color contrast and of contrast gain control. The second project aims to reveal the interaction of such local mechanisms and higher-level, perception and studies of scene segmentation. The third project aims to study the influence of global illumination on shape and color perception. Experiments in which the spatial and chromatic properties of scene lights vary are proposed to reveal central visual mechanisms that underlie shape and color constancy. The fourth project examines the detection of color patterns using visual search paradigms. The goal is to determine the visual field distribution of detection mechanisms that are sensitive to change in illumination.
